Note: Descriptions are shown in the official language in which they were submitted.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
1
COMMUNICATION IN THE PRESENCE OF UPLINK-DOWNLINK
CONFIGURATION CHANGE
Background
[0001] In a wireless communication system, downlink and uplink
transmissions of
information (control signaling or data) can be according to either a frequency
division
duplex (FDD) mode or a time division duplex (TDD) mode. In the FDD mode,
uplink
and downlink transmissions are separated in the frequency domain, by
transmitting
uplink data using a first carrier frequency, and transmitting downlink data
using a
second carrier frequency. In the TDD mode, on the other hand, both uplink and
downlink transmissions occur on the same carrier frequency; however, uplink
and
downlink transmissions are separated in the time domain, by sending uplink and
downlink transmissions in different time periods.
[0002] In some wireless communications systems, different uplink-downlink
configurations may be defined. A particular uplink-downlink configuration can
specify that, within a frame, a first subset of subframes in the frame is used
for uplink
transmissions, and a second subset of subframes in the frame is used for
downlink
transmissions. Different uplink-downlink configurations can employ different
numbers of uplink and downlink subframes.
Summarv
[0003] In general, according to some implementations, a method of a user
equipment (UE) includes receiving downlink data wirelessly from a network node
in a
first frame according to a first uplink-downlink configuration; and in
response to a
configuration change that causes a second frame following the first frame to
be
according to a second, different uplink-downlink configuration, determining a
reference uplink-downlink configuration in response to the configuration
change.
The method further includes sending, in the second frame, an acknowledgement
indication for the downlink data according to a timing of the reference uplink-
downlink configuration.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
2
[0004] In general, according to further implementations, a network node
includes a
communication interface to communicate with a user equipment (UE), and at
least
one processor. The at least one processor is configured to receive uplink data
in a
first frame according to a first uplink-downlink configuration, and in
response to a
configuration change that causes a second frame following the first frame to
be
according to a second, different uplink-downlink configuration, determine,
based on
at least one rule relating to a combination of the first and second uplink-
downlink
configurations, a timing of an acknowledgement indication in the second frame
for
acknowledging the uplink data.
[0005] In general, according to further implementations, a network node
includes a
communication interface to communicate with a user equipment (UE), and at
least
one processor. The at least one processor is configured to detect a
configuration
change from a first uplink-downlink configuration to a second uplink-downlink
configuration, and in response to the configuration change, determine, based
on at
least one rule, a timing relationship between an uplink grant and a scheduled
subframe of a frame for communicating uplink data, the at least one rule
specifying
that the timing relationship of the uplink grant is determined by an uplink-
downlink
configuration of a frame carrying the uplink grant, unless the configuration
change
involves one of an enumerated set of configuration combinations.
[0006] Other or alternative features will become apparent from the
following
description, from the drawings, and from the claims.
Brief Description Of The Drawings
[0007] Some embodiments are described with respect to the following figures:
Fig. 1 is a schematic diagram of different uplink-downlink configurations,
according to some examples;
Figs. 2-4 are schematic diagrams illustrating a change of uplink-downlink
configurations;
Fig. 5 is a flow diagram of a downlink hybrid automatic repeat request
(HARQ) operation according to some implementations;
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
3
Figs. 6-9 are schematic diagrams depicting timing relationships for
communicating an acknowledgement indication of downlink data in response to a
configuration change between uplink-downlink configurations, according to some
implementations;
Figs. 10-11 are schematic diagrams depicting uplink grant timing in the
presence of a configuration change between different uplink-downlink
configurations,
according to further implementations;
Figs. 12-16 are schematic diagrams depicting timing relationships for the
transmission of an acknowledgement indication and for data retransmission for
uplink data in the presence of a configuration change between uplink-downlink
configurations, according to further implementations;
Fig. 17 is a block diagram of the example system according to some
implementations.
Detailed Description
[0008] In a wireless communications network, different time division duplex
(TDD)
mode configurations may be employed. Such configurations can be referred to as
TDD uplink-downlink configurations (or more simply, uplink-downlink
configurations),
such as those used in a Long Term Evolution (LTE) network that operates
according
to LTE standards provided by the Third Generation Partnership Project (3GPP).
The
LTE standards are also referred to as the Evolved Universal Terrestrial Radio
Access (E-UTRA) standards. Although reference is made to LTE in the ensuing
discussion, it is noted that techniques or mechanisms according to some
implementations can be applied to other wireless access technologies.
[0009] An uplink-downlink configuration defines a number of uplink and
downlink
subframes that can be used within a frame structure. According to LTE, a frame
structure is referred to as a radio frame, where the radio frame has a number
of
subframes. A subframe refers to a segment of an overall frame, where the
segment
has a specified time interval.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
4
[0010] Fig. 1 shows an example table listing seven different uplink-
downlink
configurations that are used for TDD communications in an LTE network. A first
column 102 of the table shown in Fig. 1 identifies the seven different uplink-
downlink
configurations (0-6). A second column 104 refers to the corresponding downlink-
to-
uplink switch-point periodicity (or more simply, "switching periodicity"),
which
represents a period in which the same switching pattern is repeated between
the
uplink and the downlink. Accord ing to LTE, the switching periodicity can be 5
milliseconds (ms) or 10 ms. Uplink-downlink configuration 1 has a 5 ms
downlink-to-
uplink switch-point periodicity, for example.
[0011] As depicted in a third column 106 in the table of Fig. 1, a frame is
divided
into 10 subframes, having subframe numbers 0-9. In the table, "D" represents a
downlink subframe, "U" represents an uplink subframe, and "S" represents a
special
subframe which includes three parts: a downlink pilot time slot (DwPTS), an
uplink
pilot time slot (UpPTS), and a guard period (GP). Downlink transmissions on a
physical downlink shared channel (PDSCH) can be made in a D subframe or in the
DwPTS portion of a special subframe. The guard period (GP) of a special (S)
subframe is to provide a transition interval between switching from downlink
transmissions to uplink transmissions.
[0012] In the ensuing discussion, a "downlink" subframe can refer to either
a D
subframe or an S subframe.
[0013] As indicated in column 106, uplink-downlink configuration 2 has a 5
ms
switching periodicity. The frame for uplink-downlink configuration 2 can be
divided
into two halves, where a first half includes subframe numbers 0-4, and a
second half
includes subframe numbers 5-9. The first half-frame for uplink-downlink
configuration 2 includes the following pattern of subframes: D, S, U, D, D.
The
same pattern repeats in the second half-frame for uplink-downlink
configuration 2.
According to LTE, a frame has a length of 10 ms and each subframe has a length
of
1 ms. Since the period in which the same switching pattern (D, S, U, D, D) for
uplink-downlink configuration 2 is repeated is the period of half a frame
(five
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
subframes or 5 ms), the switching periodicity for uplink-downlink
configuration 2 is 5
ms.
[0014] As can be seen in Fig. 1, certain uplink-downlink configurations
supporta
larger number of downlink subframes than uplink subframes, while other uplink-
downlink configurations may support a larger number of uplink subframes than
downlink subframes. Uplink-downlink configuration 5 has the largest number of
downlink subframes, while uplink-downlink configuration 0 has the largest
number of
uplink subframes.
[0015] The different uplink-downlink configurations provide for flexibility
in terms of
proportional resources assignable to uplink and downlink communications within
a
given assigned frequency spectrum. The different uplink-downlink
configurations
allow for distribution of radio resources unevenly between uplink and downlink
communications. As a result, radio resources may be used more efficiently by
selecting an appropriate uplink-downlink configuration based on traffic
characteristics
and interference conditions in uplink and downlink communications.
[0016] For some applications, the proportion of uplink and downlink traffic
data
(e.g. web browsing data, voice data, etc.) may change relatively rapidly. In
an LTE
system, an uplink-downlink configuration for TDD mode communication can be
semi-
statically assigned every 640-ms. In other words, within the 640-ms time
interval,
the uplink-downlink configuration assigned for communications within a
particular cell
stays static. However, keeping the uplink-downlink configuration static for
such a
relatively long time interval may not lead to efficient use of radio
resources,
particularly when the traffic patterns of UEs within the cell are changing
relatively
rapidly.
[0017] In accordance with some implementations, dynamic TDD uplink-downlink
reconfiguration is provided, in which uplink-downlink configurations for
communications with a UE can be changed relatively frequently (more frequently
than allowed by current LTE standards).
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
6
[0018] Dynamically changing TDD uplink-downlink configurations can cause
timing issues associated with hybrid automatic repeat request (HARQ)
operations.
HARQ provides for the provision of error detection and correction information
from a
transmitter to a receiver in association with transmitted data to allow the
receiver to
detect and possibly correct errors in the data. The receiver can provide
either a
positive acknowledgement (ACK) or a negative acknowledgement (NACK) in
response to receiving the data. If the transmitter receives an ACK from the
receiver
in response to previously sent data, then the transmitter can transmit new
data to the
receiver. However, if the transmitter receives a NACK, then the HARQ process
of
the transmitter can retransmit the previously sent data.
[0019] The LTE standards provide for a timing lin kage (or timing
relationship)
between the transmission of data (either downlink data or uplink data) and the
responsive return of the acknowledgement indication (either ACK or NACK) from
the
receiver. Such timing linkage (or timing relationship) can include a downlink
HARQ
ACK/NACK timing lin kage (for acknowledging downlink data) or an uplink HARQ
ACK/NACK timing lin kage (for acknowledging uplink data).
[0020] Accord ing to LTE, the downlink HARQ operation (for acknowledging
downlink data sent from a wireless access network node to a UE) is
asynchronous,
in which the receiver does not know ahead of time what is being transmitted or
when. Stated differently, the receiver (more specifically the UE) does not
know
which HARQ process at the wireless access network node is transmitting the
downlink data, and does not know ahead of time the redundancy version (RV)
value
of the downlink transmission. Different RV values represent different
combinations
of data, error detection information, and error correction information sent
from a
transmitter to a receiver. The wireless access network node can send the HARQ
process identifier (for identifying an HARQ process) and the RV value in a
resource
allocation message that is sent on a control channel to the UE, in some
examples.
For example, if the downlink data is sent in the PDSCH, then the HARQ process
identifier can be sent in a PDSCH resource allocation message sent on a
physical
downlink control channel (PDCCH).
CA 02908650 2015-10-01
WO 2014/165350
PCT/US2014/031706
7
[0021] For
an uplink transmission, synchronous HARQ operation is performed, in
which the wireless access network node knows ahead of time the HARQ process
and RV value associated with uplink data to be transmitted by a UE within a
particular subframe. An HARQ process refers to an instance of an HARQ entity
that
controls an HARQ operation. Multiple HARQ processes can be executed in
parallel,
for communicating respective blocks of data (downlink or uplink data).
[0022] To address timing relationships impacted by TDD uplink-downlink
configuration changes, techniques or mechanisms according to some
implementations are able to send acknowledgement indications of uplink or
downlink
data in appropriate subframes.
[0023] An example of a timing linkage issue associated with a downlink HARQ
operation when a configuration change occurs from uplink-downlink
configuration 4
to uplink-downlink configuration 5 is depicted in Fig. 2.
[0024] Fig. 2 shows two frames 202 and 204, where a first frame 202 is
according
to uplink-downlink configuration 4, whereas a second frame 204 immediately
following the first frame 202 is according to uplink-downlink configuration 5.
In the
example of Fig. 2, a configuration change (also referred to as a
reconfiguration) has
occurred between frames 202 and 204. According to uplink-downlink
configuration
4, the acknowledgement indication (ACK or NACK) for a downlink data
transmission
(on PDSCH, for example) at any of subframes 6, 7, 8, and 9 are to be
transmitted on
uplink subframe 3 in the next frame, as indicated by arrows 206. In other
words,
downlink data transmission at subframe 6, 7, 8, or 9 is linked to subframe 3.
However, as depicted in Fig. 2, subframe 3 in the second frame 204 for
configuration
is a downlink subframe, rather than an uplink subframe as would be the case
for
uplink-downlink configuration 4. As a result, subframe 3 in frame 204
according to
uplink-downlink configuration 5 cannot be used to send an ACK/NACK in the
uplink
direction to acknowledge transmitted downlink data in subframe 6, 7, 8, or 9
in the
first frame 202. In the example of Fig. 2, after the uplink-downlink
configuration
change, the second frame 204 has an arrangement of subframes that is
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
8
incompatible with the timing linkage expected by the uplink-downlink
configuration
prior to the configuration change.
[0025] Fig. 3 illustrates an issue associated with transmitting an
acknowledgement
indication for uplink data in response to a TDD uplink-downlink configuration
change,
which in Fig. 3 is a change from uplink-downlink configuration 2 (for a first
frame
302) to uplink-downlink configuration 1 (for a second frame 304 immediately
following the first frame 302). According to the timing relationship specified
by
uplink-downlink configuration 2, the ACK/NACK for uplink data (such as uplink
data
sent in the physical uplink shared channel or PUSCH) at subframe 7 is supposed
to
be transmitted at downlink subframe 3, as indicated by arrow 306. However, in
the
second frame 304 after the configuration change, subframe 3 is no longer a
downlink
subframe, but rather, is an uplink subframe. Accord ingly, the ACK/NACK for
the
uplink data sent at subframe 7 in the first frame 302 cannot be sent in
subframe 3 in
the second frame 304.
[0026] As noted above, the uplink HARQ operation is synchronous. As a result,
in
addition to the uplink HARQ timing issue discussed in connection with Fig. 3,
a grant
timing relationship also has to be considered in response to an uplink-
downlink
configuration change. To enable a UE to send uplink data, a wireless access
network node sends an uplink grant to the UE, where the uplink grant
identifies the
next subframe in which the UE is to send uplink data. Note that the next
uplink data
to be sent by the UE in response to the uplink grant can be a transmission of
new
uplink data or a retransmission of previously sent uplink data that was not
received
by the wireless access network node.
[0027] Fig. 4 shows an example of a configuration change from uplink-
downlink
configuration 0 (for a first frame 402) to uplink-downlink configuration 2
(for a second
frame 404 immediately following the first frame 402). As indicated by arrow
406, the
acknowledgement indication for the uplink data sent at subframe 2 in the first
frame
402 is sent by the wireless access network node in the downlink in subframe 6
of the
first frame 402. In the example of Fig. 4, it is assumed that the
acknowledgement
indication sent in subframe 6 of the first frame 402 is NACK, which indicates
that the
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
9
uplink data sent at subframe 2 in the first frame 402 was not successfully
received by
the wireless access network node. In response to the NACK at subframe 6 in the
first frame 402, the retransmission of the uplink data should occur in
subframe 3 in
the next frame, as indicated by arrow 408. However, due to the configuration
change, subframe 3 in the next frame 404 is a downlink subframe, instead of an
uplink subframe, so that the retransmission of the uplink data cannot occur in
subframe 3 in the second frame 404. As a result, retransmission of the uplink
data
cannot occur in subframe 3 of the second frame 404, as expected by the timing
relationship for configuration 0.
[0028] Simply declaring data reception error and passing the error to upper
layers
of a protocol stack in response to failure to receive an acknowledgement (ACK
or
NACK) can lead to increased traffic delay and decreased efficiency of radio
resource
usage if configuration changes occur relatively frequently.
[0029] In accordance with some implementations, continuity of the HARQ
timing
relationship is provided for both uplink and downlink HARQ operations after a
configuration change. The following discusses examples relating to techniques
or
mechanisms provided for downlink HARQ operation in the presence of a TDD
uplink-
downlink configuration change, and an uplink HARQ operation in the presence of
a
TDD uplink-downlink configuration change.
[0030] In some implementations, it is assumed that a TDD uplink-downlink
configuration change occurs at a frame boundary (a boundary between frames). A
UE may be informed of the configuration change before the frame boundary.
Different UEs may be informed of the change at different times. In the ensuing
discussion, it is assumed that the configuration change occurs at the boundary
of
frame n and frame n-i-1.
[0031] In some specific examples, the downlink HARQ ACK/NACK timing
relationship can be according to Table 1 below, which is reproduced from Table
10.1.3.1-1 in 3GPP TS 36.213. Table 1 associates an uplink subframe n, which
conveys ACK/NACK, with downlink subframes n-kõ 1=0 to M-1.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
Table 1: Downlink association set index K: {ko,ki,= = .4_1}
Uplink-downlink Subframe n
Configuration
0 1 2 3 4 5 6 7 8 9
0 - - 6 - 4 - - 6 - 4
1 - - 7,6 4 - - -
7,6 4 -
2 - - 8, 7, 4, 6 - - - - 8, 7, 4, 6
- -
3 - - 7, 6, 11 6,5 5,4 - - - _ _
4 - - 12, 8, 7, 11 6, 5, 4, 7 - - - _ -
-
5 - - 13,12,9,8,7,5,4,11,6 - - -
- - _ _
6 - - 7 7 5 - - 7 7 -
[0032] Downlink HARQ Operation In The Presence Of A TDD Uplink-Downlink
Configuration Change
[0033] For a downlink HARQ operation, a downlink data (e.g. PDSCH)
transmission or retransmission is to be properly acknowledged after the TDD
uplink-
downlink configuration change. Fig. 5 is a flow diagram of a downlink HARQ
operation according to some implementations. The process of Fig. 5 can be
performed by a UE, in accordance with some implementations.
[0034] The process receives (at 502) downlink data in a first frame according
to a
first uplink-downlink configuration. In response to a configuration change
that
causes a second frame following the first frame to be according to a second,
different uplink-downlink configuration, the process determines (at 504) a
reference
uplink-downlink configuration based on at least one of the first and second
uplink-
downlink configurations.
[0035] The process then sends (at 506), in the second frame, an
acknowledgement indication for the received downlink data according to a
timing of
the reference uplink-downlink configuration.
[0036] Note that the determining (at 504) performed according to Fig. 5 can
include selecting an uplink-downlink configuration, based on at least one rule
(discussed below). The selected reference uplink-downlink configuration can be
the
same as the first or second uplink-downlink configuration, or can be different
from
the first and second uplink-downlink configurations.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
11
[0037] Fig. 6 illustrates an example of a configuration change from
configuration 0
(for a first frame 602) to configuration 2 (for a second frame 604 immediately
following the first frame 602). In the first frame 602, downlink data
transmitted in
each downlink subframe, except subframe 6, of the first frame 602 can be
acknowledged within the same frame 602, as indicated by arrows 606, 608, and
610.
If the timing relationship of configuration 0 for downlink HARQ operation is
followed,
then the downlink data sent in subframe 6 should be acknowledged (by
communicating an ACK or NACK) at subframe 2 in the second frame 604 (as
indicated by arrow 612). Note that subframe 2 in a frame is always an uplink
subframe in ail seven of the uplink-downlink configurations shown in Fig. 1.
Accordingly, if the configuration of the first frame 602 is configuration 0,
then the
downlink HARQ timing of configuration 0 can be followed regardless of the
configuration of the second frame 604. In the example of Fig. 6, the reference
uplink-downlink configuration is configuration 0, which is determined (at 504)
based
on a rule governing the case where the first frame before the configuration
change is
configuration 0.
[0038] In alternative implementations, the downlink HARQ timing can be
according
to configuration 5 in the second frame following the configuration change,
regardless
of the configuration of the first frame prior to the configuration change.
This is
possible because configuration 5 (as shown in Fig. 1) has just one uplink
subframe
(subframe 2), while the remaining subframes are downlink subframes. Thus, the
downlink HARQ timing of configuration 5 uses subframe 2 to acknowledge (by
sending an ACK or NACK) ail downlink data sent in an immediately preceding
frame.
In such alternative implementations, the reference uplink-downlink
configuration
determined (at 504) in Fig. 5 is configuration 5.
[0039] In further implementations, if the configuration change involves
configurations that use the 5-ms switching periodicity (configurations 0, 1,
2, and 6 in
Fig. 1), then the downlink HARQ timing can follow the timing of configuration
2
regardless of the configuration of the first and second frames before and
after the
configuration change. In such further implementations, the reference uplink-
downlink configuration is configuration 2, according to a rule governing the
case
CA 02908650 2015-10-01
WO 2014/165350
PCT/US2014/031706
12
where the configuration change involves configurations that use the 5-ms
switching
periodicity (in other words, both the first and second frames before and after
the
configuration change use the 5-ms switching periodicity).
[0040] A more generalized implementation for downlink HARQ operations that
addresses a change from any first uplink-downlink configuration to any other
uplink-
downlink configuration is provided below. In a first frame (current frame n)
prior to a
configuration change, downlink data transmission in certain subframes of the
first
frame may be acknowledged within the same first frame (current frame n).
However,
at least one downlink data transmission in the current frame n has to be
acknowledged in the next frame n-i-1 following the configuration change.
[0041] The following rules can be specified regarding downlink HARQ timing.
The
downlink HARQ timing (or more specifically, the PDSCH HARQ timing) can be
based on the configuration of the current frame n, if the downlink data can be
acknowledged within this current frame n. However, if the acknowledgement
indication (ACK or NACK) has to be provided in the subsequent frame n+1, then
the
timing follows a reference uplink-downlink configuration (determined at 504 in
Fig. 5),
which can be determined based on the set of downlink subframes of the
configuration of the current frame n and the set of subframes of the
configuration of
the subsequent frame n+1 following the configuration change. A set DLSF
(DownLink SubFrame) can represent the set of downlink subframes (which
includes
the D subframes and S subframes in Fig. 1, for example) within a specific
radio
frame. DLSF(n) represents the set of downlink subframes in frame n. DLSF(n+1)
represents the set of downlink subframes in frame n+1. In a first case (Case
1), if
DLSF(n) is a superset of DLSF(n+1), then the reference configuration is the
configuration of the current frame n, in which case the HARQ timing follows
the
timing of the current configuration (of frame n), which is specified as the
reference
configuration. DLSF(n) is a superset of DLSF(n+1) if each downlink subframe in
DLSF(n+1) is included in DLSF(n).
[0042] In a
second case (Case 2), if DLSF(n) is a subset of DLSF(n+1), then the
reference configuration is the configuration following the configuration
change (in
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
13
other words, the configuration of frame n-i-1). DLSF(n) is a subset of
DLSF(n+1) if
each downlink subframe in DLSF(n) is included in DLSF(n+1).
[0043] In a third case (Case 3), if DLSF(n) is neither a superset nor a
subset of
DLSF(n+1), then the reference configuration is the configuration that includes
downlink subframes that make up the union set of downlink subframes of the
configuration before and the configuration after the configuration change. In
other
words, the reference configuration in this case is the uplink-downlink
configuration
having a set of downlink subframes according to DLSF(n) u DLSF(n+1).
[0044] Assuming that the uplink-downlink configuration before a
configuration
change is referred to as configuration a and the uplink-downlink configuration
after
the configuration change is configuration b, then the combination of uplink-
downlink
configurations before and after the configuration change can be represented as
(a,b). In view of the foregoing, different combinations fall into the three
cases (Cases
1, 2, and 3) as follows:
Case 1(18 combinations) where DLSF(n) D DLSF(n+1): (1,0), (1,6), (2,0), (2,1),
(2,6), (3,0), (3,6), (4,0), (4,1), (4,3), (4,6), (5,0), (5,1), (5,2), (5,3),
(5,4), (5,6), and
(6,0);
Case 2 (18 combinations) where DLSF(n) c DLSF(n+1): (0,1), (0,2), (0,3),
(0,4),
(0,5), (0,6), (1,2), (1,4), (1,5), (2,5), (3,4), (3,5), (4,5), (6,1), (6,2),
(6,3), (6,4), and
(6,5);
Case 3 (6 combinations) where neither DLSF(n) D DLSF(n+1) nor DLSF(n) c
DLSF(n+1): (1,3), (3,1), (2,3), (2,4), (3,2), and (4,2).
[0045] In alternative examples, instead of using the different HARQ timings
of
Case 1, Case 2, and Case 3 set forth above, the reference configuration can be
defined according to Case 3. In other words, the reference configuration to be
used
for HARQ timing in the presence of a configuration change is the uplink-
downlink
configuration that includes the union of DLSF(n) and DLSF(n+1).
[0046] An example involving a configuration change from configuration 1 (frame
702) to configuration 3 (frame 704) is depicted in Fig. 7. Downlink data
transmissions in subframes 0, 1, and 4 in frame n (702) can be acknowledged
within
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
14
the same frame n so they follow the downlink HARQ timing of configuration 1
(see
arrows 706, 708, and 710). Downlink data transmissions in subframes 5, 6, 9 of
frame n are acknowledged in the following frame, n-i-1.
[0047] Comparing DLSF(n) with DLSF(n+1) results in a determination that Case 3
above should be used as the reference configuration. The reference
configuration is
the uplink-downlink configuration that includes a set of subframes according
to the
union of DLSF(n) and DLSF(n+1). Taking the union of the DLSF(n) and DLSF(n+1)
sets corresponding to configurations 1 and 3, respectively, results in the
reference
configuration being configuration 4. Therefore, following the downlink HARQ
timing
of configuration 4, subframe 5 in frame n is acknowledged at subframe 2 in
frame
n+1, and subframes 6 and 9 in frame n are acknowledged at subframe 3 in frame
n+1 (see arrows 712, 714, and 716).
[0048] In alternative implementations, configuration 3 can be used as a
reference
configuration, in which case subframe 9 in frame n is acknowledged with
subframe 4
in frame n+1 (arrow 718). As a further alternative, configuration 1 can be
used as
the reference configuration, in which case subframe 6 in frame n is
acknowledged at
subframe 2 in frame n+1 (arrow 720).
[0049] Table 2 below summarizes the reference configuration for cross-frame
HARQ timing of ail possible uplink-downlink configuration combinations,
including a
combination in which the original configuration is configuration 0, for use
due to a
configuration change. Multiple numbers in some entries represent alternative
possible choices. An alternative mapping is also shown in Table 2 below where
the
destination configuration is always assumed for original uplink-downlink
configuration
0 and 1 (in other words, for a configuration change where the original
configuration is
0 or 1, the reference configuration can be the destination configuration).
CA 02908650 2015-10-01
WO 2014/165350
PCT/US2014/031706
Table 2: Reference configuration for cross frame PDSCH HARQ timing
Original Uplink-Downlink Destination Uplink-Downlink Configuration after
change
Configuration
0 1 2 3 4 5 6
0 - 0 or 1 0 or 2 0 or 3 0 or 4 0 or
5 0 or 6
1 1 - 2 4 or 1 4 or 1 5 1 or
6
or 3
2 2 2 - 5 or 3 5 or 4 5 2
3 3 4 5 - 4 5 3
4 4 4 5 4 - 5 4
5 5 5 5 5 5 _ 5
6 6 1 2 3 4 5 -
[0050] In alternative implementations, configuration 5 can be used as the
reference configuration for ail reconfiguration combinations. Since, with
configuration 5, ail the PDSCH acknowledgement indications will be sent at
subframe 2 which is always an uplink subframe, the timing linkage can be
satisfied
for any uplink-downlink configuration combination.
[0051] Fig. 8 shows an example where configuration 5 is used as the reference
configuration regardless of the reconfiguration combination. In the example of
Fig. 8,
the configuration change occurs at the boundary between frames 802 and 804,
and
the configuration change is from configuration 1 to configuration 3. As
indicated by
arrows 808, 810, and 812, downlink data in subframes 0, 1, and 4 in frame 802
can
be acknowledged within the same frame 802. However, the acknowledgement
indications for subframes 5 and 6 in frame 802 are acknowledged in subframe 2
in
the next frame 804, as indicated by arrows 814 and 816.
[0052] The timing linkage for the acknowledgement indications for subframes 5
and 6 in frame 802 follow the reference configuration that is selected to be
configuration 5. However, according to configuration 5, an acknowledgement
indication for subframe 9 in frame 802 would also have to be communicated in
subframe 2 following the configuration change. However, as depicted in Fig. 8,
there
is an insufficient time length between subframe 9 in frame 802 and subframe 2
in
frame 804. In some examples, a minimum of 4 ms has to exist between the
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
16
transmission of downlink data and the subsequent communication of the
acknowledgement indication. Since subframe 2 in frame 804 following frame 802
cannot satisfy this minimum time length, a subsequent frame 806 that is after
frame
804 has to be used to commun icate the acknowledgement indication for the
downlink data in subframe 9 of frame 802. Arrow 818 indicates that the
acknowledgement indication for subframe 9 in frame 802 is transmitted in
subframe
2 in frame 806 that is the second frame following frame 802.
[0053] As noted above, if the configuration change is confined within
uplink-
downlink configurations with 5-ms switching periodicity, another alternative
is to use
the downlink HARQ timing linkage of uplink-downlink configuration 2 as the
reference configuration for ail reconfiguration combinations.
[0054] As another alternative, Fig. 9 shows an example where downlink data at
subframe 9 of frame n (902) cannot be acknowledged using subframe 3 of frame n-
i-1
(904) (arrow 906) due to the link direction change at subframe 3 in frame n+1
after
the configuration change from configuration 1 to configuration 2. In Fig. 9,
instead of
sending an acknowledgement indication in subframe 3 of frame n-i-1, the
wireless
access network node can send a small (e.g. 1 resource block) HARQ
retransmission
of the corresponding transport block after the reconfiguration boundary at
subframe
3 of frame n-i-1. Note that the retransmitted transport block in subframe 3 in
frame
n-i-1 does not actually contain the data that was originally sent in subframe
9 in frame
n. The retransmitted transport block is a shell or container that includes
information
identifying the previously sent data, without including the data.
[0055] In response to the retransmitted transport block in subframe 3 of
frame
n-i-1, the UE sends an ACK/NACK following the HARQ timing of configuration 2,.
Therefore, the UE can convey the ACK/NACK at subframe 7 of frame n-i-1, as
indicated by arrow 908. Note that the ACK/NACK at subframe 7 of frame n-i-1 is
an
acknowledgement indication for the original downlink data transmitted at
subframe 9
in frame n, as indicated by arrow 910.
[0056] As a further alternative, if there is no timing linkage for downlink
HARQ
operation due to an uplink-downlink configuration change, NACK can be assumed
by
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
17
the wireless access network node. In this case, the wireless access network
node
can send a retransmission of the correspond ing transport block afterward and
receive an ACK/NACK in response to the retransmitted transport block.
[0057] In general, according to the foregoing alternative implementations,
a
wireless access network node sends downlink data to a UE in a first frame
according
to a first uplink-downlink configuration. In response to a configuration
change that
causes a second frame following the first frame to be according to a second,
different uplink-downlink configuration, the wireless access network node
determines
that a subframe in the second frame is unavailable for receiving an
acknowledgement indication from the UE of the downlink data. In response, the
wireless access network node re-transmits a transport block corresponding to
the
downlink data to the UE, and the wireless access network node receives an
acknowledgement indication in response to the re-transmitted transport block.
[0058] Acknowledqement Bundlinq Versus Acknowledqement Multiplexinq
[0059] The acknowledgement indication (ACK or NACK) for downlink data (sent in
PDSCH) is transmitted either on the PUCCH or the PUSCH, depending upon
whether there is simultaneous uplink data to send in the same uplink subframe.
In
some cases, the ACK/NACK associated with more than one PDSCH can be mapped
into a single uplink subframe. Two downlink HARQ acknowledgement modes are
supported in TDD operation according to LTE: ACK/NACK bundling mode and
ACK/NACK multiplexing mode. In the bundling mode, a logical AND operation of
ACK/NACKs for multiple downlink subframes whose associated ACK/NACKs are
mapped into the same uplink subframe can be performed. With the bundling mode,
a single ACK/NACK (created by the logical AND operation) is transmitted in
response to multiple downlink data transmissions whose ACK/NACKs map to the
same uplink subframe.
[0060] With the multiplexing mode, the ACK/NACKs that are mapped to the same
uplink subframe are transmitted separately in the same uplink subframe. In
other
words, multiple ACK/NACK bits are communicated in the same uplink subframe
that
can be detected separately by the receiver.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
18
[0061] By using ACK/NACK bundling, fewer ACK/NACK bits have to be sent. For
example, for configuration 1, data in subframes 0 and 1 are acknowledged in
subframe 7. A single ACK/NACK that is a logical AND of the ACK/NACKs for
downlink data transmissions in subframe 0 and 1 can be sent in the uplink
subframe
7. If there is an ACK for subframe 0 and a NACK for subframe 1, then a NACK is
sent as a result of the logical AND. This implies that both the downlink
transmissions
in subframes 0 and 1 would have to be retransmitted.
[0062] Uplink-downlink configuration 5 supports the bundling mode, but not
the
multiplexing mode. Thus, the use of bundling mode or multiplexing mode is to
be
considered when performing a configuration change, according to some
implementations. If a current frame n uses the multiplexing mode, then, in
some
implementations, the wireless access network node is not allowed to perform a
configuration change to configuration 5.
[0063] In alternative implementations, if the current frame n uses the
multiplexing
mode, and the wireless access network node reconfigures to configuration 5,
then
the UE can automatically switch to using the bundling mode after the
configuration
change.
[0064] Number of HARQ Processes in Downlink HARQ Operation
[0065] The maximum number of downlink and uplink HARQ processes per serving
cell for a given UE can vary with the TDD uplink-downlink configuration.
According
to the LTE standards, the maximum number of downlink HARQ processes per
serving cell is set forth in the table below:
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
19
TDD Uplink- Maximum number of
Downlink HARQ processes
configuration
0 4
1 7
2 10
3 9
4 12
15
6 6
[0066] In the uplink, the number of HARQ processes per serving cell is set
forth in
the table below:
TDD Uplink- Number of HARQ
Downlink processes for normal
configuration HARQ operation
0 7
1 4
2 2
3 3
4 2
5 1
6 6
[0067] Whenever the uplink-downlink configuration is changed, the number of
HARQ processes is changed accordingly to match to the current configuration,
in
both the uplink and downlink directions. For downlink HARQ processes, it is
relatively easier to handle the configuration change since each HARQ process
is an
asynchronous process and each downlink grant specifies an HARQ index number
(to identify an HARQ process). When the number of HARQ processes changes to a
larger or the same number after reconfiguration, the current m downlink HARQ
buffer(s) should be able to directly transfer to the first m HARQ buffer(s) of
the
downlink HARQ processes after reconfiguration. An HARQ buffer is used to
buffer
downlink data that may potentially have to be retransmitted if a NACK is
received.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
[0068] When the number of HARQ processes changes to a smaller number due to
reconfiguration, one or more of the following schemes can be used to handle
the
configuration change.
1. The number of HARQ buffers of the previous configuration is temporarily
kept. The number of HARQ buffers is reduced to that of the new
configuration by allowing the excessive number of HARQ processes to
complete.
2. The wireless access network node uses a conservative modulation and
code scheme (MCS) to send PDSCH to make sure that the UE will receive
the PDSCH correctly and complete the HARQ transmission before the
uplink-downlink configuration change in order to make the number of HARQ
processes equal to the specified maximum number after the change.
3. As soon as the wireless access network node makes the decision to
perform a configuration change, the wireless access network node may start
to control the number of HARQ processes, prior to the reconfiguration, to
the specified maximum number of the new configuration after the change by
stopping the use of the extra downlink HARQ processes after the UE has
acknowledged their contents.
4. The wireless access network node may terminate the excessive number of
HARQ processes right before the reconfiguration. The resulting packet
errors are passed to an upper layer (e.g. Radio Link Control or RLC layer)
for handling.
[0069] Downlink HARQ Buffer Size and Rate Match ing
[0070] Rate match ing for PDSCH transmission is described in Section
5.1.4.1.2 of
3GPP TS 36.212. Rate matching uses a soft buffer, and ensures that coded bits
at
the output of a rate match ing stage will fit in the soft buffer, as defined
by the soft
buffer size. The soft buffer size refers to the maximum number of received
bits to be
stored in a buffer used for HARQ packet combining for a UE. Rate match ing
creates
an output bit stream having a target code rate.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
21
[0071] The soft buffer size fora transport block is N'IR bits, and the soft
buffer size
for the r-th code block is Alcb bits. The size Alcb for the PDSCH is obtained
as follows,
where C is the number of code blocks computed in Section 5.1.2 of 3GPP TS
36.212, and K, =3Ku is the circular buffer length for the r-th code block.
N
Ncb = minIR for downlink turbo coded transport
channels
C
where AAR is equal to:
N soft
N IR ¨ (
'MIMO = min
vM DL HARQ M )
where:
Nõft is the total number of soft channel bits (see 3GPP TS 36.306);
Km is equal to 2 if the UE is configured to receive PDSCH transmissions
based on transmission modes 3 or 4 as defined in Section 7.1 of 36.213, 1
otherwise;
MDL_HARQ is the maximum number of downlink HARQ processes as defined in
section 7 in 3GPP TS 36.213; and
Mijmjt is a constant (e.g. equal to 8).
[0072] With dynamic TDD uplink-downlink reconfiguration, MDL_HARQ is the
only
quantity which may vary from one TDD uplink-downlink configuration to another
in
the above equations. TDD uplink-downlink configurations 2, 3, 4, and 5 all
have
MDLHARQ greater than Mhm,t, so switching between two configurations from this
subset will not result in any change to N'IR and Ncb. Conversely, if at least
one of the
two TDD uplink-downlink configurations used in the dynamic TDD reconfiguration
belongs to the subset {0, 1, 6}, then NIR and as a result Ncb will change as a
result of
the dynamic TDD reconfiguration.
[0073] The parameter Ncb can determine which coded bits are actually retained
and transmitted as part of the rate matching process described at the end of
Section
5.1.4.1.2 of 3GPP TS 36.212. Two possible approaches for dealing with this
potential "resizing" of Ncb due to dynamic TDD reconfiguration are as follows.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
22
[0074] First, following a TDD reconfiguration boundary (with a first TDD
uplink-
downlink configuration used before this boundary and a second different TDD
uplink-
downlink configuration used after this boundary), both new data transmissions
and
ail HARQ retransmissions (including those for which the original first
transmission of
the transport block occurred before the reconfiguration boundary) would both
use Ncb
as calculated using MDL_HARQ for the second TDD uplink-downlink configuration.
[0075] Second, following a TDD reconfiguration boundary, new data
transmissions
and HARQ retransmissions of transport blocks originating after the
reconfiguration
boundary would use Ncb as calculated using MDL_HARQ for the second TDD uplink-
downlink reconfiguration. HARQ retransmissions of transport blocks originating
before the reconfiguration boundary would use Ncb as calculated using MDL_HARQ
for
the first TDD uplink-downlink reconfiguration.
[0076] Uplink Data Grant Timing
[0077] As noted above, in the uplink, an HARQ operation is synchronous. Thus,
in addition to uplink HARQ timing relationships, grant timing relationships
also have
to be considered in response to TDD uplink-downlink configuration changes.
[0078] In accordance with some implementations, the uplink HARQ timing
relationship and uplink HARQ grant timing relationship follow the timing
relationship
specified by the uplink-downlink configuration at the frame where the uplink
transmission and uplink grant, respectively, are transmitted, with various
exceptions
as discussed below.
[0079] For retransmission of uplink data, the timing relationship follows
that of the
uplink-downlink configuration at the frame where the NACK is transmitted.
[0080] Uplink Grant Timing
[0081] The scheduling, using uplink grants, of uplink subframes in frame n-
i-1
(which follows a configuration change after frame n) can raise issues under
certain
conditions. The default uplink grant timing linkage is determined by the
current
frame's uplink-downlink configuration. Thus, the timing linkage for the uplink
grant is
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
23
determined by the uplink-downlink configuration of the frame where the
downlink
subframe carrying the uplink grant is located.
[0082] Fig. 10 shows an example of a configuration change from
configuration 0
(for current frame 1002) to configuration 6 (for second frame 1004 immediately
following the first frame 1002). Arrows 1006 and 1008 depict grant timing
relationships for uplink grants sent in subframes 5 and 6, respectively, in
the first
frame n (1002). According to arrow 1006, an uplink grant in subframe 5 in the
current frame 1002 schedules an uplink data transmission in subframe 2 in the
second frame 1004. Similarly, the arrow 1008 specifies that an uplink grant
sent in
the downlink subframe 6 in the current frame 1002 schedules an uplink data
transmission in subframe 3 of the second frame 1004.
[0083] For configuration 0, an uplink grant sent in the downlink subframe 0
in the
second frame 1004 would be able to schedule an uplink data transmission in the
uplink subframe 4 in the second frame 1004. However, because of the
configuration
change to configuration 6, the current LTE standards would not allow for the
downlink subframe 0 in the second frame 1004 to schedule an uplink data
transmission in the uplink subframe 4. According to configuration 6, the
downlink
subframe 0 in the second frame 1004 can schedule uplink data transmission in
uplink subframe 7 (arrow 1010), and the downlink subframe 1 in the second
frame
1004 can schedule uplink data transmission in uplink subframe 8 in the second
frame 1004 (arrow 1012).
[0084] To address the issue of the inability to schedule uplink data
transmission in
uplink subframe 4 in the second frame 1004 following a configuration change, a
new
uplink grant timing relationship (represented by arrow 1014) can be used,
where an
uplink grant sent in subframe 0 of the second frame 1004 can schedule uplink
data
transmission in the uplink subframe 4 in the second frame 1004. Thus,
according to
Fig. 10, multiple uplink subframes (two PDSCH subframes) can be scheduled from
one downlink subframe (subframe 0 in the second frame 1004) by using a UL
index,
where the UL index can identify subframe 4 or subframe 7 in the second frame
1004.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
24
[0085] Fig. 11 shows another example of an uplink grant timing in the
context of a
configuration change from configuration 2 (for the first frame 1102) to
configuration 0
(for the second frame 1104). Arrows 1106, 1108, 1110,1112, and 1114 depict
grant
timing relationships specified by current LTE standards. However, due to the
configuration change, an uplink grant cannot schedule an uplink data
transmission in
subframe 3 in the second frame 1104 after the configuration change. As a
result, a
new uplink grant timing link, represented by arrow 1116, is defined for
allowing an
uplink grant sent in the downlink subframe 9 of the first frame 1102 to
schedule an
uplink data transmission in the uplink subframe 3 in the second frame 1104.
[0086] Thus, according to some implementations, the following uplink grant
timing
rule can be used for scheduling uplink transmissions in uplink subframes in
the
presence of a configuration change.
[0087] Generally, the timing linkage for uplink grant (timing between an
uplink
grant and the scheduled uplink data transmission or retransmission) is
determined
by the uplink-downlink configuration of the frame where the downlink subframe
carrying the uplink grant is located, except:
(1) where the configuration of frame n before the configuration change is
configuration 2 or 5, a new uplink grant link from subframe 9 in frame n to
schedule subframe 3 in the subsequent frame n-i-1 is to be used for ail
configuration change combinations except for combination (2, 5) (change from
configuration 2 to configuration 5) or combination (5, 2); and
(2) where the configuration of frame n+1 after the configuration change is
configuration 6, subframe 0 in frame n-i-1 is used to schedule subframe 4 and
subframe 7 in frame n-i-1 in ail change combinations where frame n-i-1 uses
configuration 6.
[0088] The rule specified above indicates that, in response to a
configuration
change, a timing relationship between an uplink grant and a scheduled subframe
of
a frame for communicating uplink data is according to the uplink-downlink
configuration of the frame carrying the uplink grant, unless the configuration
change
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
involves one of an enumerated set of configuration combinations (as specified
in
exceptions (1) and (2) above).
[0089] For exception (1) above, the new uplink grant timing linkage is from
frame n
before the configuration change to a subframe in frame n-i-1 after the
configuration
change (see arrow 1116 in Fig. 11). As a result, the UE has to know the
configuration change at least one subframe in advance.
[0090] According to exception (1), if the configuration change is from
configuration
2 to 5 or 5 to 2, no new uplink timing linkage has to be used.
[0091] For exception (2) in which the configuration change causes frame n+1
to
be according to configuration 6, subframe 0 in frame n+1 initially only carnes
uplink
grant for subframe 7. However, with the new timing linkage (arrow 1014 in Fig.
10,
for example), subframe 0 in frame n+1 has to carry uplink grants for two
uplink
subframes 4 and 7 in frame n+1. One way to do this is to automatically use an
uplink grant received in subframe 0 in frame n-i-1 for uplink transmissions in
both
subframes 4 and 7, without using uplink index bits. This should only happen at
the
TDD reconfiguration boundary. An example of the uplink grant timing with this
scenario is shown in Fig. 10.
[0092] In some specific examples, the timing relationships for uplink
grant,
ACK/NACK and transmission/retransmission are specified in Table 3 below, which
is
reproduced from Table 8.2 of 3GPP TS 36.213. Upon detection of a PDCCH with
downlink control information (DCI) format 0/4 and/or a physical HARQ indicator
channe l (PHICH) transmission in subframe n intended for the UE, the UE
adjusts the
correspond ing PUSCH transmission in sub-frame n-i-k, where k is provided in
entries
in the table below.
[0093] For TDD uplink-downlink configuration 0, if the least significant
bit of the
uplink index in the DCI format 0/4 is set to 1 in subframe n or a PH 10H is
received in
subframe n=0 or 5 in the resource correspond ing to ipHicH=1, or PH 10H is
received in
sub-frame n=1 or 6, the UE adjusts the corresponding PUSCH transmission in sub-
frame n-i-7. If, for TDD uplink-downlink configuration 0, both the most
significant bit
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
26
and least significant bit of the uplink index in the DCI format 0/4 are set in
subframe
n, the UE adjusts the corresponding PUSCH transmission in both subframes n+ k
and n+7, where k is given in Table 3 below.
Table 3: k for PUSCH transmission
TDD UL/DL subframe number n
Configuration 0 1 2 3 4 5 6 7 8 9
0 46 46
1 6 4 6 4
2 4 4
3 4 44
4 44
4
6 77 77 5
[0094] Uplink HARQ and Retransmission Timing
[0095] Since the uplink HARQ operation is synchronous, various issues may
arise
as a result of an uplink-downlink configuration change that may not be present
for
the downlink HARQ operation. In particular, when an uplink-subframe heavy
configuration is changed to a downlink-subframe heavy configuration, then some
uplink HARQ processes may not continue to proceed due to the lack of number of
uplink subframes.
[0096] The following rules can be used for uplink HARQ and retransmission
timing
during an uplink-downlink reconfiguration (a, b) from a current frame n
(having
configuration a) to the second frame n-i-1 (having configuration b).
(a) For reconfiguration combinations (0, 1), (1, 0), (1, 6), (6, 1),and (0,
6), the
ACK/NACK for uplink data (e.g. PUSCH) sent at subframe 8 in frame n before
the configuration change follows the uplink HARQ timing of the configuration
after the configuration change in frame n+1.
(b) For reconfiguration combinations (0, 2), (1, 2), (2, 0), (2, 1), (2, 6),
and (6, 2),
the ACK/NACK for uplink data sent at subframe 7 in frame n before the
configuration change follows the uplink HARQ timing of the configuration after
the configuration change in frame n-i-1.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
27
(c) For reconfiguration combinations (3, 1) and (3, 2), the ACK/NACK for
uplink
data sent at subframe 4 in frame n before the configuration change is sent at
subframes 1 and 3, respectively, in frame n-i-1.
(d) For reconfiguration combination (0,6), subframe 0 in frame n+1 conveys the
ACK/NACKs for both uplink data sent at subframes 3 and 4 in frame n before
the configuration change, and retransmissions of uplink data occurs at
subframes 4 and 7 of frame n-i-1, respectively, in response to correspond ing
NACKs.
[0097] An example of rule (a) is illustrated in Fig. 12. During the
reconfiguration,
uplink-downlink configuration 1 (for frame 1202) is changed to configuration 0
(for
frame 1204). The ACK/NACK for the uplink data at subframe 8 in frame n (1202)
is
supposed to be transmitted at subframe 4 in the following frame n-i-1 (1204).
However, subframe 4 becomes an uplink subframe in frame n-i-1 as a result of
the
configuration change. Arrow 1206 shows a new timing relationship for the
ACK/NACK for the uplink data at subframe 8 in frame n: the new timing
relationship
is according to configuration 0 (for frame n-i-1), which specifies that
subframe 5 of
frame n-i-1 is to be used for communicating the ACK/NACK for the uplink data
at
subframe 8 in frame n.
[0098] Arrows 1208, 1210, and 1212 in Fig. 12 depict timing relationships
for
ACK/NACKs of uplink data sent in subframes 2, 3, and 7 of frame n; these
timing
relationships are according to configuration 1. Arrows 1214, 1216, 1218, and
1220
depict timing relationships for retransmissions of uplink data following
respective
NACKs.
[0099] Fig. 13 shows an example of rule (b). During the reconfiguration,
uplink-
downlink configuration 2 is changed to configuration 1. The ACK/NACK for
uplink
data sent at subframe 7 in frame n (1302) is supposed to be transmitted at
subframe
3 in frame n-i-1 (1304). It is not possible since subframe 3 in frame n-i-1
has become
an uplink subframe as a result of the configuration change. Arrow 1306 shows a
new timing relationship for the uplink data sent at subframe 7 in frame n: the
new
timing relationship is according to configuration 1 (for frame n-i-1), which
specifies
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
28
that subframe 1 of frame n-i-1 is to be used for communicating the ACK/NACK
for the
uplink data at subframe 7 in frame n.
[00100] Arrow 1308 shows a timing relationship between uplink data sent at
subframe 2 in frame n and the corresponding ACK/NACK. Arrows 1310 and 1312
depict timing relationships for retransmissions of uplink data following
respective
NACKs.
[00101] An example of rule (c) is shown in Fig. 14. In this example, uplink-
downlink
configuration 3 is changed to configuration 1. Based on uplink HARQ timing of
uplink-downlink configuration 3, the ACK/NACK for subframe 4 in frame n (1402)
should be sent at subframe 0 of frame n+1. However, due to the configuration
change, there is no PHICH resource (more generally, no HARQ resource)
configured
in subframe 0 in frame n-i-1 according to configuration 1. So the new timing
relationship is represented by arrow 1406, which specifies that subframe 1 in
frame
n-i-1 is to communicate the ACK/NACK for the uplink data in subframe 4 in
frame n.
If a NACK is sent in subframe 1 in frame n-i-1, then retransmission of the
uplink data
can occur in subframe 7 in frame n-i-1 (arrow 1408), which is according to the
existing retransmission timing lin kage for configuration 1.
[00102] Arrows1410 and 1412 in Fig. 14 depict timing relationships for
ACK/NACKs
of uplink data sent in subframes 2 and 3 of frame n; these timing
relationships are
according to configuration 3. Arrows 1416 and 1418 depict timing relationships
for
retransmissions of uplink data following respective NACKs.
[00103] In rules (a)-(c), the uplink data transmission occurs in frame n,
which is
before the configuration change, and the ACK/NACK occurs in the following
frame
n-i-1. As a result, the uplink-downlink configuration of the following frame n-
i-1 does
not have to be known in advance. The location of the acknowledgement subframe
can be determined later but before the frame boundary, as soon as the
reconfiguration information becomes available.
[00104] Fig. 15 illustrates rule (d), which addresses reconfiguration from
configuration 0 to configuration 6. Due to a smaller number of downlink
subframes
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
29
with configuration 0, some of the downlink subframes are responsible for
communicating multiple ACK/NACKs and retransmissions. Following the uplink
HARQ timing of configuration 0, ACK/NACK bits for both subframes 3 and 4 in
frame
n (1502) are conveyed at subframe 0 of frame n+1 (arrows 1506 and 1508,
respectively, in Fig. 15). However, based on the uplink HARQ timing of
configuration
6, there is only one retransmission linkage from subframe 0 in frame n-i-1 to
the
subframe where the retransmission is to occur. The PHICH resource provisioned
at
subframe 0 of frame n+1 (according to configuration 6) is only for one uplink
subframe PUSCH.
[00105] One solution is to provision the PHICH resource at the UE (discussed
further below). Another solution is to combine the ACK/NACKs on the PHICH at
subframe 0 for both subframes 4 and 7 in frame n-i-1. For example, if the
wireless
access network node does not request any retransmissions (which means that the
wireless access network node has successfully received subframes 3 and 4 in
frame
n), then the wireless access network node sends an ACK at subframe 0 in frame
n-i-1, and non-adaptive retransmissions do not occur at subframes 4 and 7. If
the
wireless access network node wants a retransmission at subframe 4 and/or
subframe 7 in frame n-i-1, then the wireless access network node sends a NACK
at
subframe 0 in frame n-i-1. This NACK triggers non-adaptive retransmissions at
both
subframes 4 and 7 in frame n-i-1 (arrows 1510 and 1512, respectively).
[00106] Arrows 1514, 1516, 1518, and 1520 specify other timing relationships
from
corresponding uplink transmissions to respective ACK/NACKs. Arrows 1522, 1524,
and 1526 specify other timing relationships between NACKs and respective
retransmissions.
[00107] Fig. 16 illustrates the case of reconfiguration from configuration 0
to 2. If
the wireless access network node does not receive uplink data at subframe 2 in
frame n (1602) correctly, the wireless access network node sends a NACK at
subframe 6 in frame n (arrow 1606). In addition, the scheduler in the wireless
access network node can purposely not grant the resource for the next subframe
2 in
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
frame n-i-1 (1604), and leave the resource for the retransmission from the
NACK at
subframe 6 in frame n (arrow 1608).
[00108] Reconfiguration combination (0, 5) can follow a similar operation as
reconfiguration combination (0, 2) discussed above.
[00109] In each of the reconfiguration combinations (0, 3), (6, 3), subframes
3 and
4 in frame n both use subframe 4 in frame n+1 to send the retransmission if
requested. As a result, the scheduler in the wireless access network node has
to
consider this to avoid possible conflict.
[00110] The following describes solutions in the context of lack of PHICH
resource,
missing retransmission timing linkage, or lack of downlink subframe
availability for
ACK/NACK. In response to detecting the foregoing, one of the following actions
can
be performed to enabie the sending of an acknowiedgment indication or an
uplink
retransmission.
[00111] In some examples, adaptive uplink retransmission can be performed. In
some reconfiguration scenarios, the wireless access network node has no PHICH
resource to convey an ACK/NACK. The wireless access network node sends an
uplink grant instead with new data indicator (NDI) not toggled (which means
that the
uplink grant is requesting the retransmission of previously sent uplink data).
In
response to this uplink grant, the UE transmits the transport data in the
retransmission buffer. This technique involves the reuse of the existing DCI
format
0/4 or the use of an HARQ process ID in the uplink grant to make timing lin
kage
more flexible.
[00112] In other examples, new PHICH resources can be created. There are
several possible schemes that can be used to provision PHICH resource in a
control
reg ion. One way is to puncture the new UE PHICH resource onto the existing
control region. Another way is to use a special PDCCH with a special RNTI. The
previous PHICH process may be maintained inside the special PDCCH. The third
way is to treat ACK/NACK bits as DCI payload. Using a special Radio Network
Temporary Identifier (RNTI), channel coding, rate matching, and Quadrature
Phase
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
31
She Keying (QPSK) modulation, a channel coding rate may be semi-statically
adjustable to obtain an improved performance and capacity trade-off.
[00113] The above proposed scheme for provisioning PHICH resource in the
control region may extend to PDSCH resource region, e.g. enhanced PDCCH
(EPDCCH) region.
[00114] As alternative examples, a new timing linkage can be created. This
technique creates a new timing linkage to the nearest downlink subframe that
has
PH 10H capability and is located 4 ms after the initial PUSCH transmission.
The
ACK/NACK bit may be bundled with an existing ACK/NACK bit. For retransmission,
the new timing linkage may be created to the nearest uplink subframe which is
located 4 ms after the subframe conveying the NACK.
[00115] Alternatively, no action is performed. There will be no ACK/NACK sent
if
there is no normal PH 10H lin k. There will be no retransmission sent if there
is no
retransmission link. The wireless access network node can terminate the HARQ
process and allow an upper protocol layer (e.g. RLC layer) to handle the
resulting
packet error.
[00116] Number of Uplink HARQ Processes
[00117] The number of uplink HARQ processes also changes as the TDD uplink-
downlink configuration changes. For uplink HARQ process, it is more difficult
to
handle the transition since it is a synchronous process and the uplink grant
does not
contain the HARQ index number. However, the uplink grant has a bit used as a
new
data indicator. Similar to downlink case, when the number of HARQ processes
changes to a larger or the same number after reconfiguration, the current m
uplink
HARQ buffer(s) can directly transfer to the first m HARQ buffer(s) of the
uplink
HARQ processes after reconfiguration. Both the wireless access network node
and
UE should know the mapping of HARQ number to subframe number. When the
number of HARQ processes changes to a smaller number due to reconfiguration,
the
proposed schemes to handle the situation are as follows:
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
32
1. The uplink HARQ buffer fills up ail available uplink HARQ processes and
temporarily stores the remaining uplink HARQ buffers in the same order in a
queue at both the UE and the wireless access network node. The
remaining HARQ processes are allowed to finish one by one as soon as
there is an opportunity. For example, whenever a current HARQ completes,
the wireless access network node can send an uplink grant with a negative
new data indicator to the UE. The UE receives the grant with negative new
data indicator, and knows to restart the first HARQ process in the queue.
2. The wireless access network node indicates a conservative MCS to the UE
to send PUSCH to make sure that the wireless access network node will
receive the PUSCH correctly, and be able to complete the HARQ
transmission before the uplink-downlink configuration change in order to
make the number of HARQ processes equal to the specified number after
the change.
3. The wireless access network node may control the number of uplink HARQ
processes to the specified number after the change by purposely not
granting new data using new uplink HARQ process.
4. The wireless access network node may terminate the excessive number of
uplink HARQ processes. The resulting packet error is passed to an upper
layer for handling.
[00118] In specific examples, the uplink HARQ ACK/NACK timing linkage is
provided in Table 8.3-1 of 3GPP TS 36.213, which is reproduced as Table 4
below.
Table 4 indicates that the PH 10H ACK/NACK received in downlink subframe lis
linked with the uplink data transmission in uplink subframe i-k, with k is
given in
Table 4. In addition, for uplink-downlink configuration 0, in subframes 0 and
5, when
ipHicH=1, k=6. This is because there may be two ACK/NACKs for a UE transmitted
on the PH 10H in subframes 0 and 5, one is represented by ipHicH=1, the other
is
lPHICH=0 =
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
33
Table 4: k for HARQ ACK/NACK
Uplink-downlink subframe number i
Configuration 0 1 2 3 4 5 6 7 8 9
0 74 74
1 4 6 4 6
2 6 6
3 6 66
4 66
6
6 64 74 6
[00119] System Architecture
[00120] Fig. 17 depicts a computing system 1700, which can be any of the UE or
wireless access network node (e.g. eNB according to LTE) discussed above. The
computing system 1700 includes machine-readable instructions 1702, which are
executable on a processor (or multiple processors) 1704 to perform various
tasks
discussed above. A processor can include a microprocessor, microcontroller,
processor module or subsystem, programmable integrated circuit, programmable
gate array, or another control or computing device.
[00121] The processor(s) 1704 can be coupled to a communication interface or
component 1706 to perform communications. For example, the communication
component 1706 can perform wireless communication over an air interface, or
perform wired communication over a wired connection. In some cases, the
computing system 1700 can include multiple communication components 1706 to
communicate with respective different network nodes.
[00122] The processor(s) 1704 can also be coupled to a computer-readable or
machine-readable storage medium (or storage media) 1708, for storing data and
instructions. The storage medium or storage media 1708 can include different
forms
of memory including semiconductor memory devices such as dynamic or static
random access memories (DRAMs or SRAMs), erasable and programmable read-
only memories (EPROMs), electrically erasable and programmable read-only
memories (EEPROMs) and flash memories; magnetic disks such as fixed, floppy
and removable disks; other magnetic media including tape; optical media such
as
compact disks (CDs) or digital video disks (DVDs); or other types of storage
devices.
CA 02908650 2015-10-01
WO 2014/165350 PCT/US2014/031706
34
Note that the instructions discussed above can be provided on one computer-
readable or machine-readable storage medium, or alternatively, can be provided
on
multiple computer-readable or machine-readable storage media distributed in a
large
system having possibly plural nodes. Such computer-readable or machine-
readable
storage medium or media is (are) considered to be part of an article (or
article of
manufacture). An article or article of manufacture can refer to any
manufactured
single component or multiple components. The storage medium or media can be
located either in the machine running the machine-readable instructions, or
located
at a remote site from which mach ine-readable instructions can be downloaded
over
a network for execution.
[00123] In the foregoing description, numerous details are set forth to
provide an
understanding of the subject disclosed herein. However, implementations may be
practiced without some or ail of these details. Other implementations may
include
modifications and variations from the details discussed above. It is intended
that the
appended claims cover such modifications and variations.
